Update Speculation thread

This is a very fair point - honestly, I think having a more granular traffic selection system would help both sides of this argument.

To be fair, Cash for Clunkers "only" got rid of about 600K cars. Another major contributor to older cars going away in the US was the scrap iron price increase in the 2000s, as a result of the commodities boom. In January 2001, a ton of scrap iron was $87 - it crossed $200 in March 2004, $300 in March 2007, $400 in March 2008, and $500 in June 2008, peaking at $558 in July 2008. There were two more spikes in the price of scrap metal - January 2011 - June 2012 (a ton was worth more than $400 for that entire period, topping out at $453) and April 2021 - June 2022 (multiple months of being over $500).

However, I don't think the current traffic implies that the commodities boom and/or C4C didn't happen in the Beamverse - it probably will change in 0.39 to be more in line with reality.

There was a '92-'95 Pessima planned many years ago - however, I think it'd be better for the '88 Pessima to get a facelift and a production year extension to 1994/'95, like the sixth-gen Mitsubishi Galant. The '96 Pessima could then get a facelift (or two) to have its production run extended to 2003 or 2006 (like the eighth-gen Galant). That'd be easier to do than having two more generations of Pessima to cover those years.

Also, Beam just broke the record for monthly average increase - currently, we're at 20,771 (3,152 more average players than 0.37 (17,619)). However, one record that I'm absolutely baffled about is the Twitch peak - there were 330,413 (SteamDB) or 327,937 (Twitch) viewers in January 2025. That's way more than the usual (10-30K at most, oftentimes in the hundreds). I think that was due to Rin Penrose streaming the game - no one else I can think of would be big enough to attract that many people at once.

 

I don't like that idea at all. It has never been about doing things the traditional or easy way with Beam, so why start now? It's early access; I say accept that things will get better incrementally, however, maybe not as fast as you'd like. Continuously developing AI traffic opens the door for more possibilities, especially dynamic and unique interactions with traffic vehicles.

 

I like the traffic variety the game has now. Over time they'll add more modern-day cars, but I don't want the 80s ones to disappear from the Beam streets. Pick two from a list of say, a modern-day Covet, X3/X5 equivalent ETK crossover, modern midsize pickup truck, Explorer-equivalent 3-row crossover, Tesla-equivalent EV, ~2010 recession-era penalty box subcompact, or large modern Euro-style van, and you'd have a more realistic traffic setup. I wouldn't think you'd have to kick out the 80s cars if you add more modern variety to the rest.

 

This could be what fake travehicles in the distance could look like.

the old cars just need to become more uncommon, they are seen to often at the moment.

 

I'm just not really sure how well that will work, though. As someone who can only play with 3-5 traffic cars (and has never had much luck getting traffic pooling to work), either a minimum of 20-30% of all cars are from the 80s/90s, or none of them are. If 1 out of 5 is too many (and statistically, it is), the only remaining option is 0 out of 5. Which brings us back to the variety issue.

The one area where we agree seems to be that more, recent-ish vehicles are called for. I love my classics and weird one-offs, but the 2000s are an inexplicable void in the roster.

 

I completely agree with you, I’m just looking from an optimization perspective: I rarely have access to a PC that can actually run with traffic at a good FPS.

 

I am also on a lower end PC, since we cant spawn as many cars you can spawn them yourself. The problem is once you spawn more than 10 cars it starts to be weird to see only old vehicles.

 

Don't forget about decimals! 1/5 is 20% out of 100%, so we need maybe 0.4/5 which is about 8% of all traffic if you were to have the AI limited to five cars. That would probably be realistic in my opinion.

 

They literally DID optimize it. They gave us simplified traffic vehicles and you can customize the amount of the traffic that spawns etc. At a certain point, minimum and recommended specs need to be considered. It’s not a developer problem, it’s a consumer problem. Traffic vehicles have a jbeam and run on the same physics engine as player vehicles. There’s no getting around the extra cpu/gpu power needed

 

There are still ways to make the illusion of more traffic without any fps loss. Also they could implement simplified physics to the traffic when they are at a distance. (Simplified physics can be found in the menu)
--- Post updated ---
An idea I had for faking heavy traffic;
https://www.beamng.com/threads/idea-for-traffic-optimisation.107934/#post-1900539

 

I—I... what?

Yeah, sure, that's certainly something you can do with pure math. I even agree with the numbers you've chosen (though using both decimals and fractions at the same time is a bit cursed IMO). But would you kindly explain to me how you plan to have 40% of a car driving around on the roads?

 

James Bond did manage that just fine...

 

What I'm trying to say is I think it would be realistic for a car being spawned to have a 8% chance of being an older vehicle. if 1/5 vehicles being spawned is an older vehicle like you said, that means in theory a vehicle being spawned into the traffic would have a 20% chance of being an older vehicle. That's probably too much for a game presumably based in the same timeframe as the real world. Approximately 9-10% of cars on US roads (which is where career mode is based) are pre-2000's which are considered older vehicles. So 8% (which was just a guess mind you) would probably be realistic.

I don't really know what's wrong with using decimals and fractions together, as they pretty much go hand in hand. With that being said I'm not great at maths though, so I could be very dramatically and embarrassingly wrong. Also don't know where 40% came in either, don't think I said that.

 

It is that 0.4/5, so if a user can run 5 traffic vehicles they need to have 0.4 or 40% of a car driving on a road, you can't have the full car, I believe it is a math humor.

 

Yeah that was an oversight, my brain is not wired for this math stuff

 

Yes, but regardless of the spawn probabilities, if one of the five traffic cars you've spawned is an 80's car, that is 20% of the traffic.

Just to nip this one in the bud in case there's confusion, BeamNG's traffic system doesn't continuously spawn and remove vehicles, like other games do. It spawns a small and discrete number of them once, and then dynamically repositions and recolors them when needed. You can change the probability of an 80's car spawning initially, but once one spawns, you're stuck with it until you choose to remove it.

It's not strictly wrong per se, it's just that since decimals and fractions are both different ways to represent some portion of something, seeing them used together feels off-putting, in a hat-on-a-hat kinda way. To clarify, 0.4/5 would be represented as the pure fraction 2/25, or the decimal 0.08, each of which reads a little cleaner (though you did also use 8%, which is the cleanest-looking option).

 

Hopefully we get better systems in the future, I didn't fully realise that it worked like this, although they probably have bigger things to work on.